The error appeared again today. I’m in the middle of installations and upgrades, but here’s what I remember which I think might be important (based on the assumption that actual applications used is not a factor):
I had upgraded Firefox from 5.0 to 6.0 and was installing Winamp, which informed me that there was an update for that. So I clicked on the button to download the new version. That should have brought up Firefox and presented me with Winamp’s home page, but Firefox did not appear, although it was running - I could see it in CIS’s Firewall pane.
Eventually I gave up (this is a slow machine, but I left it long enough), and so closed everything and restarted, assuming that either Firefox 6.0 was misbehaving with CIS, or there was some issue with Firefox 6.0 and the add-on “NoScript” which I use. The aim was to go directly to Winamp’s web site after the restart.
Upon restart, the “COMODO Application Agent is not running” thing happened. Defense+ then came up with an alert about Firefox wanting to access a DNS thing - presumably it’s not in your trusted cloud list yet - which I said to allow and to send COMODO files.
I checked Computer Management’s Event Log, and there seems to be nothing in there that was unique to that startup, certainly nothing from CIS directly. I also browsed the COMODO “Program Files” and “All Users”'s “Application Data” folders, and didn’t see anything which looked like some kind of error log.
When this happens again, is there anything I can look for?
Anyway, I tried Firefox again, and it started ok, and was able to acess Winamp’s web site and download the new version of it ok. Upon then restarting again. CIS’s “not running” error was gone.
The first time this happened with me was just after it had been installed fresh. Perhaps something to do with cloud access or handling applications which have not yet been approved?
A race condition with the processes you asked us to check for, where they are having to do alot upon startup and so are not able to respond to the main application, which then thinks they’re not running?
And a suggestion…
When the diagnostics are run, clear any alerts which are going to be rechecked, and reisse the alert if the error condition is subsequently found. That would clear what seems to be happening here, where all’s fine but the error is being reported and can’t be cleared.